Monitoring The Impact of Social Factors on Crime using Machine Learning
Abstract:
Due to the population growth crime rate is increasing, which is a great challenge for the police department. Crime analysis is a systematic way of detecting and investigating patterns and crime trends. In our work, we focus to find the effect of social and economic factors on uniform crime statistics data set. Afterwards, this study introduces a machine learning technique that uses uniform crime statistics data set to measure the impact of social and economic factors on crime in the USA. In this work, we first monitor the effect of poverty and unemployment on four crime types (robbery, property snatch crime, burglary, and motor vehicle snatch). This work is based on crime, social and economic factors of the USA data set. Secondly, the effect of GDP, GDP per capita, imports, exports, urban growth, unemployment, and inflation on the total crime of the USA from 1960 to 2012 was determined. K Nearest Neighbor, Support vector regression, Multilayer Perceptron, Linear Regression, and gradient boosting have been applied, and a comparative analysis between these techniques has been performed. The MSE rate of linear regression is lower than SVR and has a high coefficient of determination value in the first experiment. Multilayer Perceptron achieved low MSE and high r square in predicting the impact of economic factors on the total crime of the USA. The experimental results conclude that unemployment affects robbery crime.
